the home of the turntable

Tube phono stage tube rush question.

the thin end of the wedge

Re: Tube phono stage tube rush question.

Postby Hanuman » 18 Apr 2012 13:24

andyr wrote:So why would you use tubes in a phono stage at all, if some tube (hybrid!) phono stages use JFETs for the first (extremely low input) gain stage, to reduce noise?

If you ever get to hear a Rhea or similar at home, Andy, you'll know why. If you already have then you should know already. It's not always about noise for some listeners and in any case, as LD points out, the surface noise of even quiet vinyl is usually way above the noise floor of any good phono stage, solid-state or tube. As I alluded to earlier, I've heard Rheas contribute to some high-volume playback and I've not noticed noise as a distraction. Looking at other parameters, tubes have some compelling features, even in a phono amplifier. The overload margin is often gigantic relative to semiconductor designs and I duly noted, in the Stereophile review of the Rhea, that John Atkinson measured the highest overload margin he ever had to that date. Add to that the definitive characteristic of tubes that many listeners prefer - their tendency to sound better when the signal is using the non-linear portion of the envelope, another way of saying that the distortion is more euphonic.

On the subject of hybrid circuits generally, Audio Research released a couple of seminal products 25 years ago (SP10 & M300) in which a conscious effort was made to combine the two component types in a way that would simultaneously use the virtues and cancel the artefacts of each type, especially in the SP10 preamplifier. Both products were hugely successful (I owned two pairs of M300s). By the gist of EveAnna Manley's justification for using the JFET in the Steelhead it would appear that Manley were thinking the same way as ARC.
Hanuman
senior member
senior member
 
Posts: 863
Images: 67
Joined: 12 Jun 2007 07:59
Location: Bangkok

Re: Tube phono stage tube rush question.

Postby andyr » 18 Apr 2012 13:48

Hanuman wrote:
If you ever get to hear a Rhea or similar at home, Andy, you'll know why. If you already have then you should know already. It's not always about noise for some listeners and in any case, as LD points out, the surface noise of even quiet vinyl is usually way above the noise floor of any good phono stage, solid-state or tube. As I alluded to earlier, I've heard Rheas contribute to some high-volume playback and I've not noticed noise as a distraction. Looking at other parameters, tubes have some compelling features, even in a phono amplifier. The overload margin is often gigantic relative to semiconductor designs and I duly noted, in the Stereophile review of the Rhea, that John Atkinson measured the highest overload margin he ever had to that date. Add to that the definitive characteristic of tubes that many listeners prefer - their tendency to sound better when the signal is using the non-linear portion of the envelope, another way of saying that the distortion is more euphonic.



No, haven't been fortunate enough to hear a Rhea, hman. :(

Re. your comment about overload margin ... what is a good overload margin, IYO?

I tested the JFET phono stage I have been building (using a sig-gen & CRO) and I stopped increasing the input when the output was 4v (with a perfect sine wave, at 1KHz).

With 5mV in (a typical MM) the phono stage produces about 250mV out, so 4v out = 80mV in ... ie. a x16 overload margin.

That seemed to me to be more than enough ... hence I stopped testing! :D

Regards,

Andy
andyr
senior member
senior member
 
Posts: 831
Joined: 13 Jan 2003 10:57
Location: Melbourne, Oz

Re: Tube phono stage tube rush question.

Postby Hanuman » 18 Apr 2012 14:29

andyr wrote:Re. your comment about overload margin ... what is a good overload margin, IYO?

I tested the JFET phono stage I have been building (using a sig-gen & CRO) and I stopped increasing the input when the output was 4v (with a perfect sine wave, at 1KHz).

Yes, that's a good and fair question. To be sure I raised that to point to possible reasons why a designer might build a tube stage not necessarily as indicator of my own priorities. Indeed it turns out that my EAR 324 (all solid state) has virtually no overload margin at very low frequencies - if this has manifested itself in any way during playback I've never been aware of it. If your output is still going up at 80mV in (assuming the THD is not off the scale) you'd have to think that that's sufficient, wouldn't you although I've read plenty of opinion on this site that recovery from overload is the really important performance factor.
Hanuman
senior member
senior member
 
Posts: 863
Images: 67
Joined: 12 Jun 2007 07:59
Location: Bangkok

Re: Tube phono stage tube rush question.

Postby pivot » 18 Apr 2012 14:31

ld wrote:Actually, noise that arises internally in valves isn't 'thermionic' or related to heat. By and large. It is related to equivalent internal resistances, which is related to parameters of the valve, which is related to geometry and spacing of electrodes. Shape of bits of bent wire, in layman's terms. Amazing, but true.

I love valves. The whole idea of manipulating current in a vacuum with bits of bent wire is so 'essential' and somewhat wonderful.


Hey LD - what do you think of the basic info here?

http://www.john-a-harper.com/tubes201/

Pretty much follows what I got in my VERY basic 200 level "electronics for hobbyists" course way back in college. Always refered to non-frequency specific random noise as "thermal noise". I recall, dimly, an experiment where we measured the noise voltage of a resistor at room temperature, cooled and heated. I am prepared to be wrong - this was a LONG time ago.

ld wrote:No dispute that various manufacturers' valves can have somewhat different noise performance. Just that must be asssociated with variation of other parameters too, it's the way it works.


Agree, usually tubes with high noise show other issues as well, microphonics, out of spec gain variance etc.
Kevin R-M

There are more things in heaven and earth, Horatio,
Than are dreamt of in your philosophy.

Hamlet Act 1
pivot
long player
long player
 
Posts: 4136
Images: 14
Joined: 27 Dec 2002 15:31
Location: Albany, NY USA

United States of America

Re: Tube phono stage tube rush question.

Postby Hanuman » 18 Apr 2012 14:42

andyr wrote:With 5mV in (a typical MM) the phono stage produces about 250mV out, so 4v out = 80mV in ... ie. a x16 overload margin.

Then again, a Lyra Delos, say, via a 24dB SUT might swing higher than 80mV although I'm only guessing - I don't know what the maximum swing of the Delos is.
Hanuman
senior member
senior member
 
Posts: 863
Images: 67
Joined: 12 Jun 2007 07:59
Location: Bangkok

Re: Tube phono stage tube rush question.

Postby goneawol » 18 Apr 2012 16:02

pivot wrote:Hey LD - what do you think of the basic info here?

http://www.john-a-harper.com/tubes201/

It's very good, IMO, pivot. The section on noise is very good. Valve specs generally quote Req, which is the noise equivalent resistance and depends just on the spacing and shape of electrodes, mostly on Gm. This relates to the noise which is hiss. This equivalent noise resistance is specced as though the resistor (which doesn't actually physically exist) is at room temperature. All noise of this form is due to random statistical motion of charge carriers in conductors, which has a set form against temperature. In this case, it's not that parts of the valve get hot which makes it any worse, the temperature of a vacuum has no meaning after all !! In this case, the resistance is 'virtual', but behaves as if it is real, including random variation and collisions of charge carriers, and hence noise.

The noise which Hanuman describes is flicker noise, and is a non-ideal, and not part of the equivalent noise resistance. I have seldom encountered it to any extent in valves, it has a physical basis in composition and construction of the cathode, and one might forsee it at very high or low anode currents. It has the same maths as random variations in flames, and friction.

The third type of noise, partition noise, doesn't happen in triodes. So not in this case.
goneawol
member
 
Posts: 4780
Images: 389
Joined: 25 Jun 2009 14:12

Previous

Return to Cartridges and Preamps


Design and Content © Vinyl Engine 2002-2014

faq | site policy | advertising